Brain damage following severe acute normovolemic hemodilution in combination with controlled hypotension in rats.

Abstract

The reduced oxygen content and perfusion pressure during acute normovolemic hemodilution (ANH) and controlled hypotension (CH) raise concerns about hypoperfusion and ischemic injury to the brain. In this study on rats, we examined the brain damage following four different degrees of ANH combined with CH.
Forty rats were randomly assigned to receive a sham operation or CH and ANH [with a hematocrit (Hct) of 30, 25, 20 or 15%]. ANH was performed after baseline physiological parameters had been monitored for 20 min; 30 min later, CH was induced using sodium nitroprusside, and the mean arterial blood pressure was maintained at 50-60 mmHg for 1 h. Rats were killed 3.5 h after hemodilution. Ultrastructural alterations in the CA1 region of the rat hippocampus were observed, and serum concentrations of S100B and neuron-specific enolase (NSE) were measured before and after ANH.
The serum S100B concentration increased significantly in the Hct 20% + CH and Hct 15% + CH groups. However, there were no significant differences in the serum levels of NSE between the groups. In the CA1 region of the rat hippocampus, marked ultrastructural alterations, such as mitochondrial denaturalization and nucleus distortion, were observed in the Hct 20% + CH and Hct 15% + CH groups.
Severe ANH (Hct < or = 20%) combined with CH may induce cerebral damage, as confirmed by marked ultrastructural alterations in the CA1 region of the rat hippocampus and significantly increased serum levels of S100B, and should be avoided.

Full text

Page 1

Acta Anaesthesiol Scand 2007; 51: 1331–1337
Printed in Singapore. All rights reserved
# 2007 The Authors
Journal compilation # 2007 Acta Anaesthesiol Scand
ACTA ANAESTHESIOLOGICA SCANDINAVICA
doi: 10.1111/j.1399-6576.2007.01448.x
Brain damage following severe acute normovolemic
hemodilution in combination with controlled
hypotension in rats
Y. L. GE1, R. LV2, W. ZHOU3, X. X. MA2, T. D. ZHONG2and M. L. DUAN4
1Department of Anesthesiology, Subei People’s Hospital, Xuzhou Medical College, Xuzhou, and Departments of2Anesthesiology3General Surgery,
Affiliated Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou, China and4Department of Anesthesiology,
Jinling Hospital, Nanjing, China
Background and aim: The reduced oxygen content and perfu-
sion pressure during acute normovolemic hemodilution (ANH)
and controlled hypotension (CH) raise concerns about hypoper-
fusion and ischemic injury to the brain. In this study on rats, we
examined the brain damage following four different degrees of
ANH combined with CH.
Methods: Forty rats were randomly assigned to receive a sham
operation or CH and ANH [with a hematocrit (Hct) of 30, 25, 20 or
15%]. ANH was performed after baseline physiological para-
meters had been monitored for 20 min; 30 min later, CH was
induced using sodium nitroprusside, and the mean arterial blood
pressure was maintained at 50–60 mmHg for1 h. Rats werekilled
3.5 h after hemodilution. Ultrastructural alterations in the CA1
region of the rat hippocampus were observed, and serum con-
centrations of S100B and neuron-specific enolase (NSE) were
measured before and after ANH.
Results: The serum S100B concentration increased significantly
in the Hct 20% þ CH and Hct 15% þ CH groups. However, there
were no significant differences in the serum levels of NSE be-
tween the groups. In the CA1 region of the rat hippocampus,
marked ultrastructural alterations, such as mitochondrial denat-
uralization and nucleus distortion, were observed in the Hct
20% þ CH and Hct 15% þ CH groups.
Conclusion: Severe ANH (Hct ? 20%) combined with CH may
induce cerebral damage, as confirmed by marked ultrastructural
alterations in the CA1 region of the rat hippocampus and sig-
nificantly increased serum levels of S100B, and should be
avoided.
Accepted for publication 21 June 2007
Key words: Brain damage; controlled hypotension; hemodi-
lution; hippocampus; neuron-specific enolase; S100B protein.
# 2007 The Authors
Journal compilation # 2007 Acta Anaesthesiol Scand
T
patients to homologous blood (1). Acute normovole-
mic hemodilution (ANH) and controlled hypoten-
sion (CH) are independently effective in decreasing
operative blood loss and the transfusion of allogeneic
blood (2, 3). Their combination may further reduce
the need for allogeneic blood transfusion (4, 5). How-
ever, the combined use of ANH and CH is still under-
utilized, because of concerns that reduced oxygen
content and perfusion pressure may lead to ischemic
and hypoperfusion injury to vital organs (6). The organ
most susceptible to hypoxia is the brain. Although an
effect of ANH combined with CH on cerebral oxy-
genation has been reported (1), their effect on direct
cerebral damage, such as ultrastructural changes, has
not been investigated in detail.
RANSFUSION-ASSOCIATED complications have re-
sulted in the need to reduce the exposure of
In recent years, protein S100B and neuron-specific
enolase (NSE), as specific biochemical markers for
the evaluation of cerebral damage, have been studied
in various clinical settings (7–9). Many studies have
demonstrated peri-operative cerebral complications,
such as delayed awakening after anesthesia, cog-
nitive dysfunction and stroke, associated with in-
creased serum levels of these markers (10–12). S100B
is an astroglial-specific protein and NSE a neuronal-
originating protein. With cerebral damage and in-
creased permeability of the blood–brain barrier, both
markers may be released into the blood and have
a higher concentration in serum. However, to date,
no data are available on the serum levels of S100B
and NSE when ANH is combined with CH.
We have conducted a study in rats to investigate
the cerebral morphological changes and serum
1331

Page 2

concentrations of brain-originating proteins after treat-
ment with four degrees of ANH combined with CH.
The results of this study may help to determine the
safe limit of the blood-sparing technique.
Methods
Animals and treatment
Adult male Sprague-Dawley rats (body weight, 350–
400 g) were obtained from Shanghai Animal Center,
Shanghai, China, and kept in accordance with the
Institutional Animal Care Committee guidelines. The
study protocols were approved by the Institutional
Animal Care Committee of Sir Run Run Shaw Hos-
pital. The animals were housed in a room with an
ambient temperature of 22 8C and a 12-h light/dark
cycle. Animals were allowed at least 7 days to ac-
climatize before the experiments. They were anes-
thetized by the intraperitoneal administration of
1250 mg of urethane (Shanghai Chemical Reagent
Co., Shanghai, China) per kilogram of body weight,
and then intubated and mechanically ventilated with
a volume-control mode [tidal volume, 8 ml/kg; fre-
quency, 60 breaths/min; fraction of inspired oxygen
(FiO2), 1.0]. From blood gas analysis (Radiometer
ALB 500, London Scientific, London, ON, Canada),
ventilation parameters were adjusted to maintain the
arterial oxygen saturation (SaO2) at greater than 95%
and the pressure of arterial carbon dioxide (PaCO2)
within the normal physiological range.
The tail vein was cannulated with a polyethylene
catheter for the intravenous administration of solu-
tions. The macrohemodynamic parameters, includ-
ing the mean arterial blood pressure (MAP), heart
rate (HR) and central venous pressure (CVP), were
measured through the left femoral artery and vein
catheterized with a microtip transducer (Abbot Lab-
oratories Ltd., Chicago, IL).
Experimental protocol
The rats were randomly divided into five treatment
groups (eight rats per group): sham operation and
four degrees of ANH [hematocrit (Hct) of 30, 25, 20
and 15%] plus CH. After monitoring of the baseline
parameters for 20 min, the rats underwent ANH,
where arterial blood from the femoral artery was
simultaneously exchanged with an equivalent vol-
ume of hydroxyethyl starch (HES 130/0.4, 6%;
medium molecular weight, low degree of substitu-
tion; Fresenius Kabi, Bad Homburg, Germany)
infused via the tail vein. To reach different Hct
targets, the volume (V) of blood to be removed was
calculated as follows (13):
V ¼ EBV??Hi? Hf
?=Hav
where EBVis the estimated blood volume (70 ml/kg),
Hiis the initial Hct, Hfis the final Hct after ANH and
Havis the mean Hct (mean of Hiand Hf). Thirty
minutes after ANH, CH was induced with the
infusion of 0.01% sodium nitroprusside (SNP) at
0.15–15 mg/kg/mintomaintainMAPat50–60 mmHg
for 1 h. After ANH and sham operation for 3.5 h, the
animals were killed by an overdose of urethane.
During the experiment, a heating pad and heating
lamp were used to maintain the temperature of rats
at about 37 8C.
Parameter measurements
MAP, HR, CVP and the body temperature (measured
viatheanus)werecontinuouslymonitoredthroughout
the experiment. MAP, HR, CVP and arterial blood gas
were recorded immediately before hemodilution (T0),
just beforeCH (T1), inthemiddle of CH (T2),at the end
of CH (T3) and 30 min after recovery from CH (T4).
Protein levels of S100B and NSE in serum
In each group, 1 ml of arterial blood was collected for
S100B and NSE measurement at baseline and at the
end of the experiment. After centrifugation (12,000 g,
4 8C) for 5 min, the serum samples were frozen at
–70 8C for further analysis. To avoid the impact of
hemodilution on serum protein levels, we also col-
lected exchanged blood. Practical serum protein con-
centrations were calculated as follows:
Cp¼ Ce?Ve=EBV þ Cf
where CPis the practical serum protein concentra-
tion, Ceis the concentration of proteins in exchanged
blood, Veis the volume of exchanged blood, EBV is
the estimated blood volume (70 ml/kg) and Cfis
the final serum protein concentration. Serum S100B
protein and NSE concentrations were analyzed by
enzyme-linked immunosorbent assay (ELISA). Val-
ues are expressed as micrograms per litre of serum.
Histological and ultrastructural studies
After the systemic circulation had been perfused with
0.9% NaCl,the skulls of the rats were openedcarefully,
the hippocampus was immediately removed and the
CA1 region was cut into standardized sections. These
were fixed for 48 h in Somogyi solution containing
paraformaldehyde, glutaraldehyde 25% and picric
acid, pH 7.4, for electron microscopy (14). The tissue
was then dehydrated and embedded in epoxy resin.
Only the ultrastructural changes in mitochondria
wereevaluatedsemi-quantitatively(15,16).Thechanges
1332
Y. L. Ge et al.

Page 3

in other cell organelles and components are simply
described. Five electron micrographs were taken of
eachbiopsy,and,ineachmicrograph,20mitochondria
were selected at random. Each mitochondrion was
graded on a scale of 0–3 (Fig. 1): 0, normal structure;
1, slight change but clearcrests; 2, edematous change;
3, accumulation of amorphous material.
Data analysis
Data are expressed as the means ? standard devia-
tion (SD), and were analyzed by SPSS 11.0. Compar-
isons between groups involved analysis of variance
(ANOVA) and Student’s t-test. Significant differences
within groups by time, as compared with baseline,
were tested by repeated measures analysis. P < 0.05
was considered to be significant.
Results
During the entire study period, all animals were
anesthetized. The total doses of SNP administered did
not differ between the treated groups, and the need for
additionaldosesofurethanetokilltheratswas similar.
Effect of ANH with CH on macrohemodynamics
and arterial blood gas parameters
MAP, HR and CVP were comparable between the
groups at baseline (Table 1). MAP and HR did not
differ at T0and T1between the groups. With ANH
and CH treatment, HR was increased at T2and T3,
and MAP was lower at T4than at baseline. Through-
out the experiment, CVP was stable. PaO2and PaCO2
were within the normal ranges at all the measured
time points (Table 2), but the pH decreased signifi-
cantly at T3and T4in the Hct 15% þ CH group.
Effect of ANH with CH on serum concentrations
of S100B and NSE
The baseline values of the serum S100B protein and
NSE concentrations were comparable between the
groups (Fig. 2). The serum S100B protein concentra-
tion increased significantly in the Hct 20% þ CH and
Hct 15% þ CH groups at the end of the experiment,
with no significant changes in NSE concentration in
any group throughout the entire study.
Effect of ANH with CH on ultrastructural
changes in the hippocampus
The Hct 20% þ CH and Hct 15% þ CH groups
showed uniform alterations in cerebral ultrastructure
induced by ANH with CH. The mean mitochondrial
scoresinthetwogroupsweresignificantlyhigherthan
those in the sham-operated group (Fig. 3). Most nuclei
appeared normal in the sham-operated, Hct 30% þ
CH and Hct 25% þ CH groups (Fig. 4). Margination
and clumping of chromatin were seen occasionally in
the Hct 20% þ CH group, but more frequently in the
Hct 15% þ CH group. The nuclear membrane was
slightly irregular in the Hct 20% þ CH group, and
became more irregular in the Hct 15% þ CH group.
The phenomenon of glial activation, such as swollen
glia assembling around the neurons, could be seen in
the Hct 20% þ CH and Hct 15% þ CH groups.
Discussion
Both ANH and CH have independently shown
efficacy as blood-saving techniques (2, 3). Conse-
quently, their combination may produce even better
conservation of blood. However, currently, a con-
firmed safe limit of ANH with CH has not been
reported, and the greater blood-sparing benefits of
the combination must be weighed against the risk of
inadequate tissue oxygenation (4). The brain is the
most susceptible organ to ischemia and hypoxia, and
so care must be taken to avoid brain damage when
this combined blood-sparing technique is used. ANH
can increase the cerebral blood flow (CBF) to com-
pensate for decreased arterial oxygen content (17),
but, in a study on dogs, ANH with CH reduced the
Fig. 1. Postulated stages of mitochondrial reaction in the CA1
region of rat hippocampus following acute normovolemic
hemodilution (ANH) combined with controlled hypotension (CH).
(A) Mitochondrial score of 0, normal structure; (B) mitochondrial
score of 1, slight change but clear crests; (C) mitochondrial score of 2,
edematous change; (D) mitochondrial score of 3, accumulation of
amorphous material.
1333
Effect of ANH and CH on brain damage

Page 4

increased CBF (6). Cerebral autoregulation maintains
CBF constant, usually to a cerebral perfusion pres-
sure of 50 mmHg (18), whereas the combination of
ANH and CH impairs the augmented CBF induced
by hemodilution (19). Thus, the security of ANH
seems to decrease on combination with CH, and the
combination may result in cerebral tissue ischemia
and hypoxia, followed by cellular damage. However,
the risk increases only when ANH involves ex-
tremely low Hct values. We aimed to investigate
the acceptable Hct range for the ANH–CH combina-
tion to help develop a technique for its safe use.
In our study, the combination of ANH and CH did
not cause serious hemodynamic instability. HR only
increased moderately during CH, which may have
been associated with the compensation mechanism
for decreased blood pressure. In addition, at T4
(recovery after CH), MAP was lower than at baseline,
Table 1
Effect of acute normovolemic hemodilution (ANH) combined with controlled hypotension (CH) on the macrohemodynamic parameters in
rats.
VariableGroupT0
T1
T2
T3
T4
MAP (mmHg)S
30%
25%
20%
15%
S
30%
25%
20%
15%
S
30%
25%
20%
15%
109 ? 9
115 ? 9
109 ? 9
105 ? 8
112 ? 10
348 ? 18
342 ? 19
350 ? 17
352 ? 38
349 ? 14
7.13 ? 0.83
6.69 ? 1.06
6.63 ? 1.07
6.75 ? 1.13
6.50 ? 0.93
110 ? 8
115 ? 6
109 ? 9
108 ? 7
114 ? 6
347 ? 15
352 ? 9
350 ? 14
355 ? 14
347 ? 11
7.25 ? 0.71
7.00 ? 0.76
6.75 ? 1.28
7.00 ? 1.07
6.65 ? 0.70
106 ? 8
56 ? 6
54 ? 5
56 ? 5
56 ? 6
342 ? 15
369 ? 12*
375 ? 13*
382 ? 22*
380 ? 20*
7.63 ? 0.52
7.00 ? 0.76
6.78 ? 1.07
6.75 ? 1.04
6.75 ? 1.05
108 ? 6
60 ? 4
57 ? 8
58 ? 5
55 ? 4
339 ? 13
365 ? 10*
378 ? 12*
379 ? 19*
377 ? 16*
7.24 ? 0.48
7.13 ? 0.54
6.70 ? 0.62
6.95 ? 0.98
6.48 ? 0.89
105 ? 7
95 ? 4*
93 ? 4*
89 ? 5*
86 ? 4*
340 ? 17
344 ? 13
341 ? 12
350 ? 17
351 ? 16
7.13 ? 0.64
6.38 ? 0.52
6.18 ? 0.83
6.35 ? 1.25
6.28 ? 1.0
HR (beats/min)
CVP (mmHg)
CVP, central venous pressure; Hct, hematocrit; HR, heart rate; MAP, mean arterial blood pressure; S, sham-operated group; T0–T4
defined in the text; 30%, Hct 30% ANH combined with CH; 25%, Hct 25% ANH combined with CH; 20%, Hct 20% ANH combined with CH;
15%, Hct 15% ANH combined with CH.
Data are means ? standard deviation (SD).
*Significant intragroup differences (P < 0.05) from baseline.
Table 2
Effect of acute normovolemic hemodilution (ANH) combined with controlled hypotension (CH) on arterial blood gas measurements in rats.
VariableGroupT0
T1
T2
T3
T4
PaO2(kPa)S
30%
25%
20%
15%
S
30%
25%
20%
15%
S
30%
25%
20%
15%
42.8 ? 2.4
43.5 ? 2.1
42.1 ? 2.8
42.8 ? 3.3
42.5 ? 2.5
5.6 ? 0.4
5.1 ? 0.4
5.5 ? 0.3
5.2 ? 0.3
5.6 ? 0.5
7.38 ? 0.05
7.39 ? 0.04
7.40 ? 0.05
7.41 ? 0.06
7.42 ? 0.05
43.9 ? 2.3
43.1 ? 2.0
42.0 ? 2.0
43.7 ? 2.8
42.4 ? 2.7
5.5 ? 0.3
5.1 ? 0.5
5.5 ? 0.4
5.2 ? 0.4
5.3 ? 0.3
7.40 ? 0.04
7.39 ? 0.03
7.41 ? 0.02
7.42 ? 0.03
7.40 ? 0.04
44.0 ? 2.1
42.5 ? 2.0
41.3 ? 1.3
42.5 ? 2.1
41.6 ? 2.7
5.2 ? 0.5
4.9 ? 0.4
5.3 ? 0.4
5.3 ? 0.7
5.2 ? 0.4
7.39 ? 0.05
7.38 ? 0.06
7.37 ? 0.05
7.35 ? 0.03
7.32 ? 0.04
44.4 ? 1.9
42.8 ? 1.7
41.6 ? 2.1
42.1 ? 2.8
41.3 ? 2.1
5.1 ? 0.4
5.2 ? 0.5
5.1 ? 0.5
5.5 ? 0.4
5.3 ? 0.5
7.40 ? 0.08
7.40 ? 0.07
7.36 ? 0.05
7.34 ? 0.06
7.28 ? 0.07*
45.5 ? 1.9
44.0 ? 1.7
43.1 ? 2.7
45.3 ? 2.8
43.7 ? 2.1
5.3 ? 0.5
5.3 ? 0.7
5.2 ? 0.5
5.2 ? 0.4
5.5 ? 0.8
7.39 ? 0.03
7.41 ? 0.06
7.39 ? 0.05
7.38 ? 0.06
7.29 ? 0.05*
PaCO2(kPa)
pH
PaCO2, pressure of arterial carbon dioxide; PaO2, pressure of arterial oxygen; S, sham-operated group; T0–T4defined in the text; 30%, Hct
30% ANH combined with CH; 25%, Hct 25% ANH combined with CH; 20%, Hct 20% ANH combined with CH; 15%, Hct 15% ANH
combined with CH.
Data are means ? standard deviation (SD).
*Significant intragroup differences (P < 0.05) from baseline.
1334
Y. L. Ge et al.

Page 5

possibly because of the decreasing effect of HES 130/
0.4 on blood volume expansion at that time, which
was 2 h after the infusion of HES. As CVP was not
significantly different between the groups, it was not
a sensitive parameter, unlike MAP and HR. PaO2and
PaCO2were both within the normal ranges at the
measured time points. PaCO2may have fallen within
the normal range because we always adjusted the
ventilation parameters according to the results of
arterial blood gas analysis. PaO2did not differ from
baseline levels during ANH and CH, but showed
a decreasing trend, especially in the Hct 15% þ CH
group, which may have resulted in the decreased pH.
In addition, tissue hypoperfusion induced by CH
may be another explanation for the low pH.
Fig. 2. Serum concentrations of S100B protein and neuron-specific
enolase (NSE).S,sham-operated group; 30%,Hct30% þ CHgroup;
25%, Hct 25% þ CH group; 20%, Hct 20% þ CH group; 15%,
Hct 15% þ CH group. Values are means ? standard deviation
(SD). *P < 0.05 vs. baseline values. #P < 0.05 vs. sham-operated
group at the end of the experiment. CH, controlled hypotension; Hct,
hematocrit.
Fig. 3. Mean mitochondrial score in the five groups after the
experiment. S, sham-operated group; 30%, Hct 30% þ CH group;
25%, Hct 25% þ CH group; 20%, Hct 20% þ CH group; 15%,
Hct 15% þ CH group. Values are means ? standard deviation (SD).
*P < 0.05 vs. sham-operated group at the end of the experiment. CH,
controlled hypotension; Hct, hematocrit.
Fig. 4. Nuclear changes in the CA1 region of rat hippocampus
following acute normovolemic hemodilution (ANH) combined with
controlled hypotension (CH). (A) Ultrastructure of representative
nucleus in the CA1 region of the hippocampus in the sham-operated
group. (B–E) Ultrastructure of representative nuclei in the CA1
region of the hippocampus in the Hct 30% þ CH, Hct 25% þ CH,
Hct 20% þ CH and Hct 15% þ CH groups, respectively. (F, G)
Glial alteration in the Hct 20% þ CH and Hct 15% þ CH groups,
respectively. Ultrastructural changes are shown with black arrows.
Hct, hematocrit.
1335
Effect of ANH and CH on brain damage